Safe probabilistic computing

Prototype driver-less vehicles requires increasing amounts of processing power and typical approaches to safety by reducing complexity are typically absent from early prototypes which focus on 'solving autonomy'. How might we adapt, improve and apply the underlying principles of safety approaches to this new world of startups, fast paced development and probabilistic computing? What are the similarities and differences between hardware failure rate predictions and safety predictions of highly complex driving systems? Would you trust your life to the outcome of the next game of go?

Jonathan Moore is the Director of Advanced Systems with exida Consulting LLC. He has over 20 years of automotive and robotics experience in systems engineering, failure mode avoidance and standards based software development. He focuses on supporting new and existing customers with their implementation of functional safety standards. He has a Bachelor of Electrical and Electronic Systems Engineering from University of York and a Master of Science Electromagnetic Compatibility from University of York. He is a Chartered Engineer and Certified Functional Safety Engineer. Recently he was accepted by FEANI and is registered as EUR ING. His interests include solving functional safety problems, finding reasonable routes to compliance for difficult systems and architectures and finding ways of enabling the safe use of technologies normally shunned by safety engineers.